Rw Setup Api Key

Setup API Key

Guide the user through obtaining a Runway API key, installing the SDK, and configuring their project for API access.

> PREREQUISITE: Run +rw-check-compatibility first to ensure the project has server-side capability.

Step 1: Create a Runway Developer Account

Direct the user to:

1. Go to https://dev.runwayml.com/
2. Create an organization (or use an existing one)
3. Navigate to Organization Settings → API Keys
4. Click Create API Key
5. Copy the key immediately — it is only shown once and cannot be recovered

Important warnings to tell the user:

• Lost keys cannot be retrieved. If lost, disable the old key and create a new one.
• API keys are organization-scoped, not user-scoped.
• You must prepay for credits before the API will work. Minimum purchase is $10 (1,000 credits at $0.01/credit). Do this at https://dev.runwayml.com/ under billing.

Step 2: Install the SDK

Node.js

npm install @runwayml/sdk

Requires Node.js 18+. The SDK includes TypeScript type definitions.

Python

pip install runwayml

Requires Python 3.8+. Includes MyPy type annotations.

Step 3: Configure the Environment Variable

The SDK automatically reads the API key from the RUNWAYML_API_SECRET environment variable.

Option A: .env file (recommended for development)

Check if the project already has a .env file. If so, append to it. If not, create one.

RUNWAYML_API_SECRET=your_api_key_here

For Node.js projects: Ensure the project loads .env files:

• Next.js, Remix, Vite — built-in .env support, no extra setup needed
• Express/Fastify/plain Node — install dotenv:

``bash npm install dotenv ` Add to the entry point: `javascript import 'dotenv/config'; ``

For Python projects: Ensure python-dotenv is installed if not using a framework with built-in support:

pip install python-dotenv

Add to the entry point:

from dotenv import load_dotenv
load_dotenv()

Option B: System environment variable

export RUNWAYML_API_SECRET=your_api_key_here

Option C: Pass directly to the client (not recommended)

// Node.js
const client = new RunwayML({ apiKey: 'your_api_key_here' });

# Python
client = RunwayML(api_key='your_api_key_here')

Warn the user: Never hardcode keys in source code. Use environment variables or a secrets manager.

Step 4: Update .gitignore

Ensure .env is in .gitignore to prevent accidentally committing the API key:

.env
.env.local
.env.*.local

Check the existing .gitignore and add the entry if it's missing.

Step 5: Verify the Setup

Suggest the user run a quick verification:

Node.js

import RunwayML from '@runwayml/sdk';

const client = new RunwayML();
// If no error is thrown, the API key is configured correctly
console.log('Runway SDK initialized successfully');

Python

from runwayml import RunwayML

client = RunwayML()
# If no error is thrown, the API key is configured correctly
print('Runway SDK initialized successfully')

Step 6: Confirm Credit Balance

Remind the user:

• The API requires prepaid credits to function
• Minimum purchase: $10 (1,000 credits)
• Purchase at: https://dev.runwayml.com/ → Billing
• They can check their balance via the API:

// Node.js - check organization info
const response = await fetch('https://api.dev.runwayml.com/v1/organization', {
  headers: {
    'Authorization': `Bearer ${process.env.RUNWAYML_API_SECRET}`,
    'X-Runway-Version': '2024-11-06'
  }
});
const org = await response.json();
console.log('Credits:', org.creditBalance);

Security Checklist

Before moving on, verify:

• [ ] API key is stored in an environment variable, not hardcoded
• [ ] .env file is in .gitignore
• [ ] API calls will only happen server-side (not in browser-executed code)
• [ ] User has purchased credits

Next Steps

Once the API key is configured, the user can proceed with integration:

• +rw-integrate-video — Video generation (text-to-video, image-to-video)
• +rw-integrate-image — Image generation
• +rw-integrate-audio — Audio generation (TTS, sound effects, voice)
• +rw-integrate-uploads — File upload for models that require image/video input

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: runwayml
• Source: runwayml/skills
• License: MIT

Install and usage instructions live in the source repository linked above.